Variable-Scale Clustering
Xuedong Gao, Ai Wang · 2018
Human naturally analyze and decide a problem from different perspectives, hierarchies, and dimensions, that is referred to as scale transformation (ST). Clustering, as one of the most effective data analysis tools, should support this ST demand. Hence, this paper focuses on the ST problem among clustering analysis especially for decision making. We define the variable-scale dataset based on the rough set theory. What's more, an algorithm of variable-scale clustering (VSC) is also proposed. A case study shows that compared to the k-modes, the clustering results of the VSC are more available and accessible to decision makers.